Peter Svenonius
Peter Svenonius |
|
Moving Right Along is a project led by Professor Peter Svenonius, financed under the Norwegian Research Council's Yngre Fremragende Forsker program, and administered under the auspices of the Center for Advanced Study in Theoretical Linguistics (CASTL). It is a five-year project (2005-2009) with the objective of investigating cross-linguistic expressions of motion and location, focusing on adpositional systems and their functional equivalents.
This is an investigation into prepositions (like to), postpositions (like Japanese made ‘up to’), particles (like away), and the way that such notions are expressed across languages. It is well-known that languages vary dramatically in the number of adpositional elements they have; even a language as closely related to English as French seems to have far fewer prepositions. Some languages are said not to have any at all. But all languages can express the thoughts that English expresses in sentences with prepositions. To circumscribe the problem, I focus on expressions of motion (as in Mary jumped into the boat) and location (cf. Mary jumped around in the boat). These are known to show rich cross-linguistic variation (even just in Germanic, for this one example type, we see German and Icelandic with a dative-accusative alternation, Swedish and Norwegian with an intonational contrast, Dutch and Afrikaans with a word-order difference, English with the difference between in and into, not to mention the Scandinavian alternation inn – inne, all corresponding to the difference between direction and location).
The primary goals of the project are to achieve a solid characterization of some of the limits on this type of construction, i.e. so-called linguistic universals, and to achieve a solid characterization of the cross-linguistic properties of the category P.
Activities are organized around a weekly seminar each semester, in which faculty, post-doctoral researchers, and graduate students meet to discuss the syntax of adpositions and related constructions. In the course of the project, over fifty languages have come under investigation, including Chickasaw, Shua, Indonesian, Hungarian, Korean, Chinese, Japanese, Gbe, Thai, Farsi, Spanish, and Germanic and Slavic languages, among others.
We have participated in some important conferences on adpositions, including one in Venice in November 2005 and one in Utrecht in June 2006, and an upcoming one in Tromsø in May 2009. We have also organized several lecture series, including one in Amsterdam in January 2006, one in Paris in September 2006, one at the LSA Summer Institue at Stanford in 2007, in Hyderabad and Delhi in November and December 2007, and in the Basque Country in 2008.
See the Schedule page for more information.
1. Why adpositions are important
2. My interest in this work, and qualifications for it
3.1. First period: Large sample, Data collection
3.2. Second period: Detailed analysis
3.3. The final period: Focus on results
The project studies prepositions and postpositions (known collectively as adpositions) and related expressions across languages. This is an area which has been extensively studied, and therefore there is a solid foundation of previous work to build on and we have been able to start the investigation at a sophisticated level.
On the other hand, there is no consensus in the field on some fundamental matters, such as whether there is a basic grammatical category of adposition. Thus we can hope to contribute in an important way to our understanding of these constructions.
There are several reasons that prepositions and related constructions are intrinsically worth studying, independently of their history as research objects.
One reason is that predicate–argument structure lies at the core of language, and there is extensive work on verbal argument structures. A detailed study of prepositional argument structure can build on that work and help to refine it. For example, the basic asymmetry between Agent and Patient arguments (by which Agents are universally prominent over Patients in argument structure, even in radically ergative languages like Dyirbal, cf. Dixon 1972, 1994) is very well established by now. I have suggested (Svenonius 1994) that the argument structure of adpositions is similarly asymmetrical, with Figure universally dominating Ground (cf. Talmy 2000). It would be valuable to know whether this can be maintained.
Another reason to study adpositions is that, in the larger context of human cognition, a great deal of work has been done on how humans orient themselves spatially and on how they locate objects in space. Therefore we have some foundation for understanding what kinds of extralinguistic factors may have influenced the development of human expressions of space and motion. This means that claims about general cognitive structures might have some hope of being separated from claims about specifically linguistic structures (cf. Scholtz & Pullum 2001).
A third reason is that this area, expressions of motion and location, appears to be a point of rather great cross-linguistic variation. Just to take a single example, consider the way we tend to locate objects in space in Indo-European languages in terms of a speaker-oriented grid: left, right, above, below, in front of, behind. The grid can also be oriented around the hearer: The mosquito is to your left, or around some other object: The mosquito is to John’s left (this is subject to constraints imposed by the grammar; notice the contrast with The mosquito is to the left of John, which retains the speaker’s or hearer’s orientation, using John as a landmark). This spatial grid has been thought by many to be a linguistic universal, but languages spoken in Central America and in Australia have been demonstrated to use for primary orientation some geographical feature such as a coast, a river, or even the path of the sun. The foreignness of such a system might be partly appreciated by imagining that your ‘east shoe’ turns into your ‘west shoe’ when you turn around. If that sounds confusing, consider first how many people seem to have trouble with left and right, and then that Tzeltal speakers who use the geography of their region for their basic orientation do not even have general words for left and right (cf. Brown & Levinson 1993, Levinson 1996).
That is not the extent of variation, by far. Part of the problem in exploring the limits of variation in this area is that it is difficult to define the area itself. Given that the categorial status of adpositions is problematic, one cannot simply go from grammar to grammar looking at how adpositions work. First, they must be identified. In fact, it is notoriously difficult to identify them even in English: Is regarding a preposition? Is near? Is out, in out of? Is there a noun in in front of? Or in in light of? Is the word in part of the expression instead of?
Perhaps partly for this reason, there is an astonishing lack of consensus in formal linguistics today regarding the status of adpositions. Virtually every imaginable position is held seriously by some reputable linguist.
This has been true for a long time: Ross (1967) and Fillmore (1968) treated prepositions essentially as grammatical features (like case), Lakoff (1968) tried to derive them from verbs, while Klima (1965), Emonds (1972), and van Riemsdijk (1978) counted them as categories with their own syntax. But even among those taking prepositions as syntactic entities, there has never been consensus regarding important matters. While Jackendoff (1977) argued that they were somewhat like verbs and somewhat like adjectives (hence opposed to nouns), others argued that they were somewhat like nouns and somewhat like verbs (hence opposed to adjectives); Emonds (1985) argued that they should be unified with Complementizers.
This heterogeneity of positions extends into the present: some assume, like Jackendoff, that P is a lexical category which projects a rich functional structure, much like a verb (Koopman 1996, 2000, den Dikken 2002) but others still maintain, like Ross, that P is not a lexical category at all (Baker 2002, in the context of arguing that the categories V and N and A are universals), but only a functional one (Grimshaw 1991), even expressible as nominal inflection (Yadroff 2000). In fact in some sense the theoretical positions have multiplied; though the usual assumption has always been that P forms a constituent with a noun phrase complement, it has recently been argued that at least some P’s are base-generated outside VP (Kayne 2001, Cinque 2002), combining with their surface complements through movement; and even that P’s are base-generated inside DP (Pesetsky & Torrego 2002), moving out of it in the course of a derivation.
The semantic literature also shows a wide variety of approaches. Cresswell (1978) uses a traditional predicate logic, introducing contextually bound arguments in order to capture shifts in points of view. Gawron (1986a, b) treats prepositions as predicates in a situation semantics framework, postulating special rules of combination to merge predicates, allowing a verb and a preposition to share an argument. Jackendoff (1990) favors a decomposition into primitive predicates and types, coindexing arguments in various ways. Zwarts (2002) and Winter & Zwarts (2003) pursue a vector semantics approach, arguing that other treatments cannot correctly capture the entailments observed.
Prepositions and expression of location and motion have attracted the greatest attention in the cognitive grammar literature. There, the most influential contributions have been those of Langacker (1987, 1991, 1999) and Talmy (1978, 1985, 2000) (see also Svorou 1993 and Heine 1997). Talmy, for example, examining the cross-linguistic expressions of directed motion, characterizes differences in terms of conflation patterns. In an English-type language (including, for example, Chinese), Talmy notes, a verb may relatively easily be combined with an expression of a path of motion in another phrase e.g. float can combine with into the cave in The bottle floated into the cave. In a Spanish-type language, this is not typically possible; instead, verbs express path as well as motion (path and motion are ‘conflated’, in Talmy’s terms): La botella entró a la cueva (flotando) ‘the bottle entered the cave (floating).’
Talmy’s observations about this and other cases of conflation have inspired a huge volume of work on diverse languages, and there, too, many different positions are held. For example, there are differences regarding the limits of conflation, whether such things are parametric, and where the notion of directed motion comes from in an expression like Mary jumped in the boat (since jump is not inherently directional, i.e. one can jump in place, without changing location, and neither is in).
The cognitive grammar tradition of studying expressions of motion draws on the substantial body of work, noted in §1 above, on the non-linguistic perception of related concepts, and thereby provides a dimension often lacking from formalist work focusing on word order and formative cooccurrences.
Finally, the Construction Grammar work should be mentioned, e.g. Goldberg (1995, inter alia), since it has also generated a substantial body of highly relevant work, especially connected to metaphorical extensions of directed motion expressions, e.g. in resultatives and constructions like He burped his way down the hall.
It is clear that a significant work on prepositions and related expressions has a chance of communicating with researchers in all these subdisciplines. The kind of project I outline here will be of great interest not only to the formalists whose great disarray I have noted above but also to the cognitive grammarians and construction grammarians who work partly in a separate tradition.
I have been working on issues related to this problem, on and off, for over ten years. My first foray into the subject was my exploration of the verb-particle construction in English (cf. Svenonius 1992c). I expanded this investigation to the verb-particle construction in Scandinavian in my dissertation (Svenonius 1994a) and a few later papers (Svenonius 1996a, 1996b, 2003, Svenonius & Ramchand 2002). I moved into separable prefix constructions in German, Dutch, and related languages (Svenonius in press b). Although all of the publications just mentioned have been focused on Germanic languages, I have constantly been aware of and vitally interested in the cross-linguistic expression of the kinds of concepts expressed in Germanic by prepositions, particles, and separable prefixes. I have established myself at this point as a respected authority in the field on Germanic prepositions and particles and am often asked to review work on those topics (cf. my reviews, Svenonius 1996c, 2002h, to appear c).
My ultimate goal has long been an understanding of the category P cross-linguistically. Strategically, I felt that my next detailed investigation should be of a language not too distant from Germanic. It was natural, then, to move into Slavic (also because of the excellent Slavic linguistic community in Tromsø). What Germanic expresses with particles or separable prefixes, Slavic languages very often express with prefixes. I have been working with Russian, Czech, Polish, Serbian, and Bulgarian graduate students in Tromsø for the past year and a half to get a detailed sense for the Slavic prefixal system and to understand exactly in what way it is similar to or different from the Germanic particle system. I have presented this work this year at conferences in Leipzig and Oxford where it has generated some excitement; I have presented it at invited lectures at Edinburgh, Oslo, Oxford, Tromsø, Lublin, Novi Sad, and Geneva, receiving very positive feedback. The cooperation with Novi Sad was financed by a grant from NRC for a project comparing Germanic particles with Slavic prefixes. The results appear in Nordlyd (Tromsø University’s Working Papers in Language and Linguistics).
At the same time as I have been doing detailed work on Germanic and Slavic, I have been examining many languages to get a sense of the nature of the variation, and have presented some of these results in Delhi (2003), Konstanz (2002), Lund (2002), and Tromsø. Some of my preliminary results led me to the conclusion that a deeper understanding of the nature of case was necessary, something which led me to spend most of 2001 researching case (some results published in Svenonius 2001a, 2002b, 2002g; lectures delivered in Montréal, Trondheim, Berlin, Utrecht, Reykjavík, Nis, and Tromsø).
Some difficult fieldwork may become necessary in the course of the project, possibly including work with speakers of underdocumented languages.
The project will study expressions of motion and location. It is already known (e.g. from Talmy’s work) that these tend to be expressed either as in English, with prepositional phrases, or as in Spanish or Korean, with path notions expressed in the verb, or as in Russian, with verbal affixes (in Russian, these are typically doubled by prepositions). There are many mixed cases, and there are other types as well, including serial verb constructions (e.g. Yorùbá), extensive local case inventories (e.g. Lezgian), and others (cf. Talmy).
For this reason, the object of study is defined here using cognitive notions, e.g. ‘expressions of motion.’ I will focus on directed motion (I rolled the ball, I threw the ball) with and without path specification (I threw the ball in the window) and the contrast between such constructions and expressions of location (cf. the ambiguous I threw the ball in the house) and will investigate how similar thoughts are expressed in different languages. Various techniques are possible, including the use of pictures (e.g. as Slobin’s ‘frog stories’).
The first two years of the project will be spent collecting information about a wide variety of languages. Sources will be the available literature (already quite substantial), typologists and other people who are familiar with a variety of languages, and grammars. I intend to be abroad during some of the collection of this information. In Tromsø we have linguists who are native speakers of many Germanic, Romance, and Slavic languages, plus Northern Sámi, Finnish, Kitharaka (Bantu), Japanese, Korean, Chinese, Farsi, and others, and in addition we have linguists who do research on Scottish Gaelic, Bengali, Austronesian, and other languages.
I aim for a large sample of languages from all over the globe. At the same time as I collect information on the expressions of motion and so on in the different languages, I will also have to collect and organize other information about them. This is because a great number of linguistic universals have been discovered by observing correlations between seemingly unrelated parts of the grammar, as with Greenberg’s (1963) observations that OV languages tend to have case-marking, or that they tend to be postpositional, or that V-initial languages tend to place the genitive after the noun. In keeping track of this kind of information I will be aided by the work of typologists, including the recently published Oxford Atlas of the World’s Languages.
On the basis of this general information regarding a large number of languages, the various claims made in the theoretical literature will be evaluated and tested. Hypotheses can be formulated and reformulated underway, and the analytic process will interact with the data collection. Languages which seem especially problematic for otherwise strong generalizations may require closer examination. In many cases these may turn out to be obscure languages for which expert sources are not available, and primary fieldwork may become necessary.
The larger sample will be the basis for some initial hypotheses, and I expect to be able to produce publishable work on universal properties of adpositional constructions within a year or so. On the basis of the work of the first two years, I will select a smaller group of languages for more detailed study, perhaps twelve, for detailed study. These languages will be carefully chosen against a range of criteria.
One, they must be languages to which I have very good access, either because of cooperation with expert linguists with access to native speakers, or because of direct contact with reliable native speakers. Ideally, there will also be a solid body of written work on the selected languages, but not necessarily.
Two, they must represent a typologically, genetically, and geographically diverse sample. At present, I expect that such a sample would include both prepositional and postpositional languages, languages with local case and without, languages with preposition stranding and without, languages with serial verbs and without, and so on. However, the work of the first two years may suggest a different basis for the small sample.
After the initial stage, the small sample will have been selected and detailed investigation can begin. The selected languages will be examined in close detail to test and refine the hypotheses made on the basis of the large sample. If I have selected twelve languages, then I can spend an average of two months on each one over a two-year period. I will be assisted in this by my research team, which will consist at least of one Master’s student, one doctoral student, and one post-doctoral researcher, plus whoever I can recruit from the local linguistics milieu. For some languages it may be necessary for all or part of the team to travel abroad.
The actual hypotheses to be tested will be developed on the basis of the work done in the first two-year period, but at this time it might be expected that they would include such questions as the following:
Is the primary difference between V and P the presence vs. absence of tense? (Cf. Langacker’s 1987 suggestion that prepositions profile atemporal processes, while verbs profile temporal ones)
Is case assignment (i.e. the licensing of noun phrase complements) essentially the same for V and P, or essentially different? (Cf. accounts which take P to be a case-assigner, like V, vs. those which take P to be part of the extended projection of N)
Are prepositions and postpositions fundamentally alike, or not? (Cf. recent suggestions that there are no true postpositions, most apparent postpositions actually being nouns, or case markers)
When languages express Path meanings in the verb, as on Talmy’s analysis of Spanish, can this be analyzed as some kind of abstract incorporation of a Path head into the verb, as I have suggested recently for Russian, or does this kind of decompositional analysis fail when the prefixes are not overt?
These questions are given here by way of illustration. In fact, these particular questions will drive some of the work in the first two-year period, so that the second two-year period will be able to work with much more refined questions.
In the final year, the doctoral student will be writing a dissertation and I will be putting together a book detailing the results of this five-year sojourn into the human brain. This final period will be focused on results and dissemination.
Dikken, Marcel den. 2002. Review of Particle Verbs and Local Domains by Jochen Zeller. Journal of Comparative Germanic Linguistics 4:145-169.
Gawron, Jean-Mark. 1986a. Situations and prepositions. Linguistics and Philosophy 9:327-382.
Gawron, Jean-Mark. 1986b. Types, contents and semantic objects. Linguistics and Philosophy 9:427-476.
Hout, Angeliek van. 1998. Event Semantics of Verb Frame Alternations: A Case Study of Dutch and Its Acquisition. Garland, New York.
Koopman, Hilda. 2000. Prepositions, postpositions, circumpositions, and particles. In The Syntax of Specifiers and Heads, Hilda Koopman (ed.), 204-260. Routledge, London.
Langacker, Ronald. 1987. Foundations of Cognitive Grammar. Stanford.
Riemsdijk, Henk van. 1978. A Case Study in Syntactic Markedness: The Binding Nature of Prepositional Phrases. Foris, Dordrecht.
Riemsdijk, Henk van. 1990. Functional prepositions. In Unity in Diversity, H. Pinkster and I. Genée (eds.), 229-241. Foris, Dordrecht.
Rooryck, Johan. 1996. Prepositions and Minimalist case marking. In Studies in Comparative Germanic Syntax, Höskuldur Thráinsson, Samuel David Epstein and Steve Peter (eds.), 226-256. Kluwer, Dordrecht.
Svenonius, Peter. In Press. Adpositions, Particles, and the Arguments they Introduce. Argument Structure, edited by Eric Reuland, Tanmoy Bhattacharya, and Giorgos Spathas, 71-110. John Benjamins, Amsterdam.
Svenonius, Peter. To appear. Spatial P in English. Cartography of Syntactic Structures, vol. 6, edited by Guglielmo Cinque and Luigi Rizzi. Oxford University Press, Oxford.
Svenonius, Peter. 2006. The Emergence of Axial Parts. Nordlyd: Tromsø Working Papers in Language and Linguistics 33.1: 49-77.
Svenonius, Peter. 2003. ‘Limits on P: Filling in holes vs. falling in holes,’ in Proceedings of the 19th Scandinavian Conference of Linguistics. Nordlyd: Tromsø Working Papers in Language and Linguistics 31.2: 431-445.
Svenonius, Peter. 2003. ‘Swedish particles and directional prepositions,’ in Grammar in Focus: Festschrift for Christer Platzack (vol. II), ed. by Lars-Olof Delsing, Cecilia Falk, Gunnlög Josefsson, and Halldór Á. Sigurdsson, Department of Scandinavian Languages, Lund, pp. 343–352.
Svenonius, Peter. 2002b ‘Icelandic Case and the Structure of Events’ Journal of Comparative Germanic Linguistics 5:197-225.
Svenonius, Peter. 2002c ‘The lexical syntax and lexical semantics of the verb-particle construction’ (with Gillian Ramchand), in Proceedings of WCCFL 21 , 387-400, Cascadilla Press.
Svenonius, Peter. 2002g, ‘Case is uninterpretable aspect,’ in Proceedings of Perspectives on Aspect Conference at Utrecht (www.uilots.lets.uu.nl/conferences/Perspectives_on_Aspect/ P_o_A_procpapers.html)
Svenonius, Peter. 2002h. Review of Zeller (2001) Particle Verbs and Local Domains. Linguist List 13:743.
Svenonius, Peter. 2001a. ‘Case and Event Structure.’ in ZASPIL 26 (Zentrum für allgemeine Sprachwissenschaft Papers in Linguistics 26), ed. by Nina Zhang (www.zas.gwz-berlin.de/papers/zaspil/infos/).
Svenonius, Peter. 1996a ‘The Verb Particle Alternation in English and Scandinavian,’ ms. University of Tromsø.
Svenonius, Peter. 1996b, ‘The optionality of particle shift,’ Working Papers in Scandinavian Syntax 57:47-75.
Svenonius, Peter. 1996c, Review of Particles by M. den Dikken, in Language 74: 816-820
Svenonius, Peter. 1994a, Dependent Nexus: Subordinate Predication Structures in English and the Scandinavian Languages , PhD dissertation, UCSC.
Svenonius, Peter. 1992c, ‘Movement of P0 in the English Verb-Particle Construction,’ in A. Black and J. McCloskey (eds.), Syntax at Santa Cruz , vol. 1 pp. 93-113. Syntax Research Center, Santa Cruz, Ca.
Talmy, Leonard. 1978. Figure and Ground in Complex Sentences. In Universals of Human Language, Joseph H. Greenberg (ed.), 625-649. Stanford University Press, Stanford, Ca.
Talmy, Leonard. 1985. Lexicalization Patterns: Semantic Structure in Lexical Forms. In Language Typology and Syntactic Description, I: Clause Structure, Timothy Shopen (ed.), 57-149. Cambridge University Press, Cambridge.
Talmy, Leonard. 2000. Toward a Cognitive Semantics, vol I: Concept Structuring Systems. MIT Press, Cambridge, Ma.
Zeller, Jochen. 2001a. Particle Verbs and Local Domains. John Benjamins, Amsterdam.
Zeller, Jochen. 2001b. Lexical particles, semi-lexical postpositions. In Semi-Lexical Categories: The Function of Content Words and the Content of Function Words, Norbert Corver and Henk van Riemsdijk (eds.), 505-549. Mouton de Gruyter, Berlin.